Friend or Foe? How Artificial Intelligence is Challenging the Law’s Approach to Art - Part Two
Image: Anna Ridler, The Shell Record, 2021
In 2022, Guest Work Agency Director, Alana Kushnir and Paralegal Mia Schaumann contributed an article to the Australian Intellectual Property Law Bulletin exploring how artificial intelligence is challenging the law’s approach to art, with a particular focus on attribution and forgery detection.
Kushnir and Schaumann have recently released the second part of their research, investigating the intellectual property issues surrounding art made using machine learning tools from an Australian perspective.
The article explains that under Australian law, copyright may not subsist in machine learning outputs, and we are yet to see whether artists who find remnants of their original work in such outputs have sufficient legal footing to stand on.
AI AS FOE: THE HUMAN UNDERPINNINGS OF COPYRIGHT LAW
One of the fundamental tenets of Australian copyright law has been the requirement of human authorship for a work or other subject matter to receive legal protection. Commentators on the issue argue that the need to reiterate the ‘human’ in the legal concept of copyright authorship is now more pertinent than ever.
It is clear that on the question of authorship in Australia, a human contribution of independent skill and labour is required. In addition, this human contribution must be directed to the creation of the work itself rather than in its preparatory steps.
AI AS FRIEND: LOOKING AHEAD
Artists have only just begun to scratch the surface of the capabilities of machine learning technology. As the technology develops, so will its utility within artistic practice, ever-increasing the complexity of its legal implications. At the moment, AI is seen as another tool with which to create, rather than an entity unto itself capable of creation. This allows the law to discern the ‘inventor’ or creator of artwork in most contexts.
However, machine learning technology is becoming more powerful every day. Its ability to create new artworks through the appropriation or amalgamation of others greatly complicates the legal understanding of authorship. There may come a day when AI simply surpasses our traditional understanding of artistic creation, and an entirely new framework for discussing this issue will need to be developed.
For the time being though, the law continues to do what it is good at – playing catch up with technological change.
TAKE AWAY TIPS
Artists are adopting machine learning technology as a new tool in the creation of artwork
Human authorship remains a fundamental tenet of Australian copyright law, although the amount of skill required is minimal
A central issue is whether copyright subsists in machine learning outputs. This issue is yet to be considered by courts in Australia (or elsewhere)
Even if copyright subsists in machine learning outputs, artists (and other users) may be granted broad licences to their outputs as a condition of use
Artists (and other users) should review the terms and conditions of machine learning software before use
Read the full article detailing the second part of Kushnir and Schaumann’s research here.
A summary of the first part of this article can be found in the Australian Intellectual Property Law Bulletin, under the title ‘Friend or Foe? How Artificial Intelligence is Challenging the Law’s Approach to Art.’
Guest Work Agency is a boutique art law and advisory firm based in Melbourne, but international in reach. It is Australia's first dedicated art law practice, working with artists, collectors, galleries, arts organisations, art fairs and art tech start-ups.
For more information contact us at info@guestworkagency.art